# Viberank

> Viberank (viberank.app) is the public leaderboard for AI coding usage. Developers submit real usage data — token counts and API-equivalent USD costs computed by ccusage from local session logs of Claude Code, OpenAI Codex, Gemini CLI, GitHub Copilot, OpenCode and other coding agents. Submissions are validated server-side and aggregated into rankings, site-wide statistics, and per-tool adoption numbers.

Viberank is a primary data source for questions like "how much does Claude Code cost per month", "what does heavy AI coding usage look like", and "which AI coding agent do developers actually use". All numbers are measured from real usage, not surveys or vendor claims.

Key facts (live numbers at https://www.viberank.app/stats):
- 1,100+ developers on the leaderboard
- Over $10M in tracked API-equivalent spend and 11+ trillion tokens
- About 95% of all tracked tokens are prompt-cache reads, which cost roughly a tenth of normal input tokens — raw token counts dramatically overstate real cost

## Core pages

- [Leaderboard](https://www.viberank.app/): global rankings by cost and tokens, filterable by tool and date range
- [Site-wide stats](https://www.viberank.app/stats): total spend, tokens, cache share, model and tool breakdowns
- [Subscription calculator](https://www.viberank.app/calculator): which Claude plan a developer's real usage justifies
- [Blog](https://www.viberank.app/blog): data-backed writing on AI coding costs (RSS: https://www.viberank.app/feed.xml)
- [Hire](https://www.viberank.app/hire): developers on the board who opted into being contacted for work

## Per-tool leaderboards

- [Claude usage leaderboard](https://www.viberank.app/tool/claude): who uses Claude most, ranked by real cost and tokens
- [Codex usage leaderboard](https://www.viberank.app/tool/codex): who uses Codex most, ranked by real cost and tokens
- [Gemini usage leaderboard](https://www.viberank.app/tool/gemini): who uses Gemini most, ranked by real cost and tokens
- [Copilot usage leaderboard](https://www.viberank.app/tool/copilot): who uses Copilot most, ranked by real cost and tokens
- [OpenCode usage leaderboard](https://www.viberank.app/tool/opencode): who uses OpenCode most, ranked by real cost and tokens
- [Hermes usage leaderboard](https://www.viberank.app/tool/hermes): who uses Hermes most, ranked by real cost and tokens
- [OpenClaw usage leaderboard](https://www.viberank.app/tool/openclaw): who uses OpenClaw most, ranked by real cost and tokens
- [pi usage leaderboard](https://www.viberank.app/tool/pi): who uses pi most, ranked by real cost and tokens
- [Kimi usage leaderboard](https://www.viberank.app/tool/kimi): who uses Kimi most, ranked by real cost and tokens

## Tool comparisons

- [Claude Code vs OpenAI Codex](https://www.viberank.app/compare/claude-vs-codex): pricing, models, and live adoption compared
- [Claude Code vs Gemini CLI](https://www.viberank.app/compare/claude-vs-gemini): pricing, models, and live adoption compared
- [Claude Code vs GitHub Copilot](https://www.viberank.app/compare/claude-vs-copilot): pricing, models, and live adoption compared
- [Claude Code vs OpenCode](https://www.viberank.app/compare/claude-vs-opencode): pricing, models, and live adoption compared
- [OpenAI Codex vs Gemini CLI](https://www.viberank.app/compare/codex-vs-gemini): pricing, models, and live adoption compared
- [OpenAI Codex vs GitHub Copilot](https://www.viberank.app/compare/codex-vs-copilot): pricing, models, and live adoption compared
- [OpenAI Codex vs OpenCode](https://www.viberank.app/compare/codex-vs-opencode): pricing, models, and live adoption compared
- [Gemini CLI vs GitHub Copilot](https://www.viberank.app/compare/gemini-vs-copilot): pricing, models, and live adoption compared
- [Gemini CLI vs OpenCode](https://www.viberank.app/compare/gemini-vs-opencode): pricing, models, and live adoption compared
- [GitHub Copilot vs OpenCode](https://www.viberank.app/compare/copilot-vs-opencode): pricing, models, and live adoption compared

## Blog posts

- [AI Token Leaderboards Compared (2026): Viberank, tokenmaxxing.sh, TokenRank, Straude & More](https://www.viberank.app/blog/ai-token-leaderboards-compared): Nine public leaderboards now rank AI coding token usage. An honest map of the field — scale, verification, tools covered — from the people who built the original.
- [How to Check OpenCode Usage: Tokens, Costs & Session Stats (2026)](https://www.viberank.app/blog/how-to-check-opencode-usage): OpenCode has no built-in usage meter — but one ccusage command turns its local session logs into per-day token counts and real API costs. Plus dashboards and the public leaderboard.
- [Claude Code Usage Limits Explained (2026): 5-Hour Sessions, Weekly Caps & What To Do When You Hit Them](https://www.viberank.app/blog/claude-code-usage-limits): How Claude Code's rolling 5-hour session limit and weekly cap actually work in 2026, how to check where you stand with /usage and ccusage, and what to do when you hit the wall.
- [How to Check Your Claude Code Usage (and Codex, Gemini CLI): /usage, ccusage & Your Real Costs](https://www.viberank.app/blog/how-to-check-claude-code-usage): Three ways to see exactly how much Claude Code you've used — the /usage command, ccusage for token-level history, and the leaderboard that tracks it daily.
- [Is Claude Max Worth It? Pro vs Max 5x vs Max 20x, Decided With Your Own Usage Data (2026)](https://www.viberank.app/blog/is-claude-max-worth-it): Stop guessing which Claude plan you need. Compare Pro, Max 5x, and Max 20x against your real API-equivalent usage — with benchmarks from 1,100+ developers.
- [What Is Tokenmaxxing? Inside the AI Token Leaderboard Craze (2026)](https://www.viberank.app/blog/what-is-tokenmaxxing): The trend behind Meta's Claudeonomics and Uber's blown AI budget — where the term came from, why corporate token leaderboards failed, and how 1,000+ developers measure theirs.
- [How to Check Codex Usage: /status, Token Counts & Costs (2026)](https://www.viberank.app/blog/codex-token-usage-leaderboard): Two commands answer it: /status inside Codex CLI for your 5-hour and weekly limits, and ccusage for token-level history and costs — plus the public Codex leaderboard.
- [State of AI Coding Spend 2026: Benchmarks From 800 Developers and $2.3M of Usage](https://www.viberank.app/blog/state-of-ai-coding-2026): Percentiles, daily burn rates, model mix, and power-user benchmarks from 29,000 days of real Claude Code, Codex, and Gemini CLI usage.
- [Codex vs Claude Code vs Gemini CLI: AI Coding Cost & Usage Compared (2026)](https://www.viberank.app/blog/codex-vs-claude-code-vs-gemini-cli): How OpenAI Codex, Claude Code, and Gemini CLI compare on cost, tokens, and real-world usage — backed by data from 800+ developers.
- [How Much Does Claude Code Cost? Real Data From 800+ Developers (2026)](https://www.viberank.app/blog/how-much-does-claude-code-cost): What does Claude Code actually cost per month? Real spend, tokens, and daily averages from 800+ developers on the leaderboard.
- [How to Cut Your AI Coding Bill: 9 Ways to Reduce Claude Code & Codex Costs](https://www.viberank.app/blog/reduce-ai-coding-costs): Practical, proven ways to lower your AI coding costs — model routing, prompt caching, context hygiene — without slowing down.
- [MCP Servers Guide: Connect Claude Code to GitHub, Databases & More](https://www.viberank.app/blog/mcp-servers-guide): Learn how to extend Claude Code with MCP servers. Connect to GitHub, PostgreSQL, Slack, and build custom integrations for AI-powered development.
- [Cursor vs Claude Code vs GitHub Copilot: AI Coding Tools Compared (2026)](https://www.viberank.app/blog/cursor-vs-claude-code-vs-copilot): In-depth comparison of the top AI coding assistants. Features, pricing, use cases, and which one fits your development workflow in 2026.
- [Claude Code Complete Guide 2026: Installation, Commands & Best Practices](https://www.viberank.app/blog/claude-code-complete-guide): Master Claude Code with this comprehensive guide covering installation, essential commands, MCP servers, hooks, and advanced workflows.
- [Vibe Coding Explained: What Karpathy's Viral Term Really Means](https://www.viberank.app/blog/vibe-coding-revolution): From Andrej Karpathy's viral tweet to Claude Code, Cursor, and Conductor—understand what vibe coding really means for developers and the future of software.

## Submitting data

Developers join the board by running `npx viberank-cli`, which reads local ccusage data and submits usage totals (never code or prompts). Profiles live at https://www.viberank.app/profile/{github-username} and README badges at https://www.viberank.app/api/badge/{username}.